28 Development of a Hardware Benchmark for Forensic Face Detection Applications

Face detection techniques are valuable in the forensic investigation since they help criminal investigators to identify victims/offenders in child sexual exploitation material. Deep learning approaches proved successful in these tasks, but their high computational requirements make them unsuitable i...

Descripción completa

Detalles Bibliográficos
Autores: Velasco Mata, Javier, Chaves, Deisy, Mata, Verónica de, Al-Nabki, Mhd Wesam, Fidalgo, Eduardo, Alegre, Enrique, Azzopardi, George
Tipo de recurso: capítulo de libro
Fecha de publicación:2021
País:España
Institución:Universidad de Castilla-La Mancha
Repositorio:RUIdeRA. Repositorio Institucional de la UCLM
OAI Identifier:oai:ruidera.uclm.es:10578/28635
Acceso en línea:http://doi.org/10.18239/jornadas_2021.34.28
http://hdl.handle.net/10578/28635
Access Level:acceso abierto
Palabra clave:Face Detection
Benchmark
GPU
CPU
Descripción
Sumario:Face detection techniques are valuable in the forensic investigation since they help criminal investigators to identify victims/offenders in child sexual exploitation material. Deep learning approaches proved successful in these tasks, but their high computational requirements make them unsuitable if there are time constraints. To cope with this problem, we use a resizing strategy over three face detection techniques —MTCNN, PyramidBox and DSFD— to improve their speed over samples selected from the WIDER Face and UFDD datasets across several CPUs and GPUs. The best speed-detection trade-off was achieved reducing the images to 50% of their original size and then applying DSFD. The fastest hardware for this purpose was a Nvidia GPU based on the Turing architecture.